🔀 Merge pull request #662 from ruby/backport/v0.5/raw_data-warnings

🔒 Fix CRLF injection vulnerabilities (backports #657, #658, #659, #660, #636, #661)

Author: nevans

lib/net/imap.rb

@@ -462,6 +462,9 @@ module Net
   #   +LITERAL-+, and +SPECIAL-USE+.</em>
   #
   # ==== RFC2087: +QUOTA+
+  # +NOTE:+ Only the +STORAGE+ quota resource type is currently supported.
+  # - Obsoleted by <tt>QUOTA=RES-*</tt> [RFC9208[https://www.rfc-editor.org/rfc/rfc9208]],
+  #   although the commands are backward compatible.
   # - #getquota: returns the resource usage and limits for a quota root
   # - #getquotaroot: returns the list of quota roots for a mailbox, as well as
   #   their resource usage and limits.
@@ -578,6 +581,16 @@ module Net
   #   See FetchData#emailid and FetchData#emailid.
   # - Updates #status with support for the +MAILBOXID+ status attribute.
   #
+  # ==== RFC9208: <tt>QUOTA=RES-*</tt>
+  # +NOTE:+ Only the +STORAGE+ quota resource type is currently supported.
+  # - Obsoletes the +QUOTA+ [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]]
+  #   extension and provides strict semantics for different resource types.
+  # - #getquota: returns the resource usage and limits for a quota root
+  # - #getquotaroot: returns the list of quota roots for a mailbox, as well as
+  #   their resource usage and limits.
+  # - #setquota: sets the resource limits for a given quota root.
+  # - Updates #status with <tt>"DELETED"</tt> and +DELETED-STORAGE+ attributes.
+  #
   # ==== RFC9394: +PARTIAL+
   # - Updates #search, #uid_search with the +PARTIAL+ return option which adds
   #   ESearchResult#partial return data.
@@ -698,13 +711,12 @@ module Net
   #
   # === \IMAP Extensions
   #
-  # [QUOTA[https://www.rfc-editor.org/rfc/rfc9208]]::
-  #   Melnikov, A., "IMAP QUOTA Extension", RFC 9208, DOI 10.17487/RFC9208,
-  #   March 2022, <https://www.rfc-editor.org/info/rfc9208>.
+  # [QUOTA[https://www.rfc-editor.org/rfc/rfc2087]]::
+  #   Myers, J., "IMAP4 QUOTA extension", RFC 2087, DOI 10.17487/RFC2087,
+  #   January 1997, <https://www.rfc-editor.org/info/rfc2087>.
   #
-  #   <em>Note: obsoletes</em>
-  #   RFC-2087[https://www.rfc-editor.org/rfc/rfc2087]<em> (January 1997)</em>.
-  #   <em>Net::IMAP does not fully support the RFC9208 updates yet.</em>
+  #   *NOTE*: _obsoleted_ by RFC9208[https://www.rfc-editor.org/rfc/rfc9208]
+  #   (March 2022).
   # [IDLE[https://www.rfc-editor.org/rfc/rfc2177]]::
   #   Leiba, B., "IMAP4 IDLE command", RFC 2177, DOI 10.17487/RFC2177,
   #   June 1997, <https://www.rfc-editor.org/info/rfc2177>.
@@ -756,6 +768,11 @@ module Net
   #   Gondwana, B., Ed., "IMAP Extension for Object Identifiers",
   #   RFC 8474, DOI 10.17487/RFC8474, September 2018,
   #   <https://www.rfc-editor.org/info/rfc8474>.
+  # [{QUOTA=RES-*}[https://www.rfc-editor.org/rfc/rfc9208]]::
+  #   Melnikov, A., "IMAP QUOTA Extension", RFC 9208, DOI 10.17487/RFC9208,
+  #   March 2022, <https://www.rfc-editor.org/info/rfc9208>.
+  #
+  #   Obsoletes RFC2087[https://www.rfc-editor.org/rfc/rfc2087].
   # [PARTIAL[https://www.rfc-editor.org/info/rfc9394]]::
   #   Melnikov, A., Achuthan, A., Nagulakonda, V., and L. Alves,
   #   "IMAP PARTIAL Extension for Paged SEARCH and FETCH", RFC 9394,
@@ -769,6 +786,7 @@ module Net
   #
   # === IANA registries
   # * {IMAP Capabilities}[http://www.iana.org/assignments/imap4-capabilities]
+  #   * {IMAP Quota Resource Types}[http://www.iana.org/assignments/imap4-capabilities#imap-capabilities-2]
   # * {IMAP Response Codes}[https://www.iana.org/assignments/imap-response-codes/imap-response-codes.xhtml]
   # * {IMAP Mailbox Name Attributes}[https://www.iana.org/assignments/imap-mailbox-name-attributes/imap-mailbox-name-attributes.xhtml]
   # * {IMAP and JMAP Keywords}[https://www.iana.org/assignments/imap-jmap-keywords/imap-jmap-keywords.xhtml]
@@ -779,8 +797,8 @@ module Net
   # * {GSSAPI/Kerberos/SASL Service Names}[https://www.iana.org/assignments/gssapi-service-names/gssapi-service-names.xhtml]:
   #   +imap+
   # * {Character sets}[https://www.iana.org/assignments/character-sets/character-sets.xhtml]
+  #
   # ==== For currently unsupported features:
-  # * {IMAP Quota Resource Types}[http://www.iana.org/assignments/imap4-capabilities#imap-capabilities-2]
   # * {LIST-EXTENDED options and responses}[https://www.iana.org/assignments/imap-list-extended/imap-list-extended.xhtml]
   # * {IMAP METADATA Server Entry and Mailbox Entry Registries}[https://www.iana.org/assignments/imap-metadata/imap-metadata.xhtml]
   # * {IMAP ANNOTATE Extension Entries and Attributes}[https://www.iana.org/assignments/imap-annotate-extension/imap-annotate-extension.xhtml]
@@ -1828,12 +1846,18 @@ def xlist(refname, mailbox)
     # to both admin and user.  If this mailbox exists, it returns an array
     # containing objects of type MailboxQuotaRoot and MailboxQuota.
     #
+    # *NOTE:* Currently, Net::IMAP only supports +QUOTA+ responses with a single
+    # resource type.  This is usually +STORAGE+, but you may need to verify this
+    # with UntaggedResponse#raw_data.
+    #
     # Related: #getquota, #setquota, MailboxQuotaRoot, MailboxQuota
     #
     # ==== Capabilities
     #
-    # The server's capabilities must include +QUOTA+
-    # [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]].
+    # Requires +QUOTA+ [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]]
+    # capability, or a capability prefixed with <tt>QUOTA=RES-*</tt>
+    # {[RFC9208]}[https://www.rfc-editor.org/rfc/rfc9208] for each supported
+    # resource type.
     def getquotaroot(mailbox)
       synchronize do
         send_command("GETQUOTAROOT", mailbox)
@@ -1845,41 +1869,59 @@ def getquotaroot(mailbox)
     end
 
     # Sends a {GETQUOTA command [RFC2087 §4.2]}[https://www.rfc-editor.org/rfc/rfc2087#section-4.2]
-    # along with specified +mailbox+.  If this mailbox exists, then an array
-    # containing a MailboxQuota object is returned.  This command is generally
-    # only available to server admin.
+    # for the +quota_root+.  If this quota root exists, then an array
+    # containing a MailboxQuota object is returned.
+    #
+    # The names of quota roots that are applicable to a particular mailbox can
+    # be discovered with #getquotaroot.
+    #
+    # *NOTE:* Currently, Net::IMAP only supports +QUOTA+ responses with a single
+    # resource type.  This is usually +STORAGE+, but you may need to verify this
+    # with UntaggedResponse#raw_data.
     #
     # Related: #getquotaroot, #setquota, MailboxQuota
     #
     # ==== Capabilities
     #
-    # The server's capabilities must include +QUOTA+
-    # [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]].
-    def getquota(mailbox)
+    # Requires +QUOTA+ [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]]
+    # capability, or a capability prefixed with <tt>QUOTA=RES-*</tt>
+    # {[RFC9208]}[https://www.rfc-editor.org/rfc/rfc9208] for each supported
+    # resource type.
+    def getquota(quota_root)
       synchronize do
-        send_command("GETQUOTA", mailbox)
+        send_command("GETQUOTA", quota_root)
         clear_responses("QUOTA")
       end
     end
 
     # Sends a {SETQUOTA command [RFC2087 §4.1]}[https://www.rfc-editor.org/rfc/rfc2087#section-4.1]
-    # along with the specified +mailbox+ and +quota+.  If +quota+ is nil, then
-    # +quota+ will be unset for that mailbox.  Typically one needs to be logged
-    # in as a server admin for this to work.
+    # along with the specified +quota_root+ and +storage_limit+.  If
+    # +storage_limit+ is +nil+, resource limits are unset for that quota root.
+    # If +storage_limit+ is a number, it sets the +STORAGE+ resource limit.
+    #
+    #   imap.setquota "#user/alice", 100
+    #   imap.getquota "#user/alice"
+    #   # => [#<struct Net::IMAP::MailboxQuota mailbox="#user/alice" usage=54 quota=100>]
+    #
+    # Typically one needs to be logged in as a server admin for this to work.
+    #
+    # *NOTE:* Currently, Net::IMAP only supports setting +STORAGE+ quota limits.
     #
     # Related: #getquota, #getquotaroot
     #
     # ==== Capabilities
     #
-    # The server's capabilities must include +QUOTA+
-    # [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]].
-    def setquota(mailbox, quota)
-      if quota.nil?
-        data = '()'
+    # Requires +QUOTA+ [RFC2087[https://www.rfc-editor.org/rfc/rfc2087]]
+    # capability, or a capability prefixed with <tt>QUOTA=RES-*</tt>
+    # {[RFC9208]}[https://www.rfc-editor.org/rfc/rfc9208] for each supported
+    # resource type.
+    def setquota(quota_root, storage_limit)
+      if storage_limit.nil?
+        list = []
       else
-        data = '(STORAGE ' + quota.to_s + ')'
+        list = ["STORAGE", Integer(storage_limit)]
       end
-      send_command("SETQUOTA", mailbox, RawData.new(data))
+      send_command("SETQUOTA", quota_root, list)
     end
 
     # Sends a {SETACL command [RFC4314 §3.1]}[https://www.rfc-editor.org/rfc/rfc4314#section-3.1]
@@ -1986,7 +2028,10 @@ def lsub(refname, mailbox)
     # <tt>STATUS=SIZE</tt>
     # {[RFC8483]}[https://www.rfc-editor.org/rfc/rfc8483.html].
     #
-    # +DELETED+ requires the server's capabilities to include +IMAP4rev2+.
+    # +DELETED+ must be supported when the server's capabilities includes
+    # +IMAP4rev2+.
+    # or <tt>QUOTA=RES-MESSAGES</tt>
+    # {[RFC9208]}[https://www.rfc-editor.org/rfc/rfc9208.html].
     #
     # +HIGHESTMODSEQ+ requires the server's capabilities to include +CONDSTORE+
     # {[RFC7162]}[https://www.rfc-editor.org/rfc/rfc7162.html].
@@ -2267,11 +2312,11 @@ def uid_expunge(uid_set)
     #     Encoded as an \IMAP date (see ::encode_date).
     #
     # [When +criteria+ is a String]
-    #   +criteria+ will be sent directly to the server <em>without any
-    #   validation or encoding</em>.
+    #   +criteria+ will be sent to the server <em>with minimal validation and no
+    #   encoding or formatting</em>.
     #
-    #   <em>*WARNING:* This is vulnerable to injection attacks when external
-    #   inputs are used.</em>
+    #   <em>*WARNING:* Although CRLF is prohibited, this is vulnerable to other
+    #   types of attribute injection attack if unvetted user input is used.</em>
     #
     # ==== Supported return options
     #
@@ -2592,6 +2637,13 @@ def uid_search(...)
     #
     # +attr+ is a list of attributes to fetch; see FetchStruct documentation for
     # a list of supported attributes.
+    # >>>
+    #   When +attr+ is a String, it will be sent <em>with minimal validation and
+    #   no encoding or formatting</em>.  When +attr+ is an Array, each String in
+    #   +attr+ will be sent this way.
+    #
+    #   <em>*WARNING:* Although CRLF is prohibited, this is vulnerable to other
+    #   types of attribute injection attack if unvetted user input is used.</em>
     #
     # +changedsince+ is an optional integer mod-sequence.  It limits results to
     # messages with a mod-sequence greater than +changedsince+.
@@ -3712,7 +3764,7 @@ def fetch_internal(cmd, set, attr, mod = nil, partial: nil, changedsince: nil)
     end
 
     def store_internal(cmd, set, attr, flags, unchangedsince: nil)
-      attr = RawData.new(attr) if attr.instance_of?(String)
+      attr = Atom.new(attr) if attr.instance_of?(String)
       args = [SequenceSet.new(set)]
       args << ["UNCHANGEDSINCE", Integer(unchangedsince)] if unchangedsince
       args << attr << flags

lib/net/imap/command_data.rb

@@ -28,6 +28,7 @@ def validate_data(data)
         end
       when Time, Date, DateTime
       when Symbol
+        Flag.validate(data)
       else
         data.validate
       end
@@ -48,7 +49,7 @@ def send_data(data, tag = nil)
       when Date
         send_date_data(data)
       when Symbol
-        send_symbol_data(data)
+        Flag[data].send_data(self, tag)
       else
         data.send_data(self, tag)
       end
@@ -132,11 +133,13 @@ def send_list_data(list, tag = nil)
     def send_date_data(date) put_string Net::IMAP.encode_date(date) end
     def send_time_data(time) put_string Net::IMAP.encode_time(time) end
 
-    def send_symbol_data(symbol)
-      put_string("\\" + symbol.to_s)
-    end
-
     CommandData = Data.define(:data) do # :nodoc:
+      def self.validate(...)
+        data = new(...)
+        data.validate
+        data
+      end
+
       def send_data(imap, tag)
         raise NoMethodError, "#{self.class} must implement #{__method__}"
       end
@@ -145,15 +148,109 @@ def validate
       end
     end
 
+    # Represents IMAP +text+ data, which may contain any 7-bit ASCII character,
+    # except for +NULL+, +CR+, or +LF+.  +text+ is extended to allow any
+    # multibyte +UTF-8+ character when either +UTF8=ACCEPT+ or +IMAP4rev2+ have
+    # been enabled, or when the server supports only +IMAP4rev2+ and not earlier
+    # IMAP revisions, or when the server advertises +UTF8=ONLY+.
+    #
+    # NOTE: The current implementation does not validate whether the connection
+    # currently supports UTF-8.  Future versions may change.
+    #
+    # The string's bytes must be valid ASCII or valid UTF-8.  The string's
+    # reported encoding is ignored, but the string is _not_ transcoded.
+    class RawText < CommandData # :nodoc:
+      def initialize(data:)
+        data = String(data.to_str)
+        data = if data.encoding in Encoding::ASCII | Encoding::UTF_8
+          -data
+        elsif data.ascii_only?
+          -(data.dup.force_encoding("ASCII"))
+        else
+          -(data.dup.force_encoding("UTF-8"))
+        end
+        super
+        validate
+      end
+
+      def validate
+        if data.include?("\0")
+          raise DataFormatError, "NULL byte must be binary literal encoded"
+        elsif !data.valid_encoding?
+          raise DataFormatError, "invalid UTF-8 must be literal encoded"
+        elsif /[\r\n]/.match?(data)
+          raise DataFormatError, "CR and LF bytes must be literal encoded"
+        end
+      end
+
+      def ascii_only? = data.ascii_only?
+
+      def send_data(imap, tag) = imap.__send__(:put_string, data)
+    end
+
     class RawData < CommandData # :nodoc:
-      def send_data(imap, tag)
-        imap.__send__(:put_string, data)
+      def initialize(data:)
+        data = split_parts(data)
+        super
+        validate
+      end
+
+      def send_data(imap, tag) = data.each do _1.send_data(imap, tag) end
+
+      def validate
+        return unless data.last in RawText(data: text)
+        if text.rindex(/~?\{[1-9]\d*\+?\}\z/n)
+          raise DataFormatError, "RawData cannot end with literal continuation"
+        end
+      end
+
+      private
+
+      def split_parts(data)
+        data = data.b # dups and ensures BINARY encoding
+        parts = []
+        while data.match(/(~)?\{(0|[1-9]\d*)(\+)?\}\r\n/n)
+          text, binary, bytesize, non_sync, data = $`, !!$1, $2, !!$3, $'
+          bytesize = Integer bytesize, 10
+          parts << RawText[text] unless text.empty?
+          parts << extract_literal(data, binary:, bytesize:, non_sync:)
+          data[0, bytesize] = ""
+        end
+        parts << RawText[data] unless data.empty?
+        parts
+      end
+
+      def extract_literal(data, binary:, bytesize:, non_sync:)
+        if data.bytesize < bytesize
+          raise DataFormatError, "Too few bytes in string for literal, " \
+            "expected: %s, remaining: %s" % [bytesize, data.bytesize]
+        end
+        literal = data.byteslice(0, bytesize)
+        (binary ? Literal8 : Literal).new(data: literal, non_sync:)
       end
     end
 
     class Atom < CommandData # :nodoc:
+      def initialize(**)
+        super
+        validate
+      end
+
+      def validate
+        data.to_s.ascii_only? \
+          or raise DataFormatError, "#{self.class} must be ASCII only"
+        data.match?(ResponseParser::Patterns::ATOM_SPECIALS) \
+          and raise DataFormatError, "#{self.class} must not contain atom-specials"
+      end
+
+      def send_data(imap, tag)
+        imap.__send__(:put_string, data.to_s)
+      end
+    end
+
+    class Flag < Atom # :nodoc:
       def send_data(imap, tag)
-        imap.__send__(:put_string, data)
+        imap.__send__(:put_string, "\\#{data}")
       end
     end